Designing and testing industrial devices for 5G private networks

Article By : Jessy Cavazos

The unique capabilities of industrial devices, their integration requirements, and growing complexity create many challenges that drive the need for testing and channel modeling.

Fifth-generation wireless (5G) and multi-access edge computing (MEC) are essential technologies for tomorrow’s smart factories. The real-time aspects of Release 16—the latest 5G standard release by the 3rd Generation Partnership Project (3GPP)—have spurred the development of the Industrial Internet of Things (IIoT). Complementing 5G to support IIoT control applications, MEC provides real-time awareness to the 5G system while helping to achieve lower latency by moving compute resources closer to the user.

Upcoming standard releases will bring new capabilities for industrial applications. Release 17, which will reach commercialization in the first half of 2022, will increase the integration of 5G with time-sensitive networks (TSN). The release will also introduce features that improve 5G positioning and reduce latency, capabilities that are critical for factory automation and remote-control applications.

Next, touted as a step function increase in the evolution of the 5G standard, Release 18 will be the first release from the 3GPP officially falling under the “5G Advanced” umbrella. This release will feature major enhancements for network intelligence, including the implementation of machine-learning (ML) techniques at different levels in the network. That’s because artificial intelligence (AI) enhancements will be essential for more demanding industrial use cases.

Figure 1 Release 16 enables 5G networks to support industrial applications. Source: Analog Devices

The rise of 5G private networks

Wireless connectivity, edge compute capabilities, and AI are key enablers for the fourth industrial revolution. These capabilities bolster the adoption of 5G in the manufacturing sector. In a survey conducted by the Capgemini Research Institute with senior executives from 1,000 industrial organizations recently, 30% said they were at the piloting stage or beyond and 40% said they expected to roll out 5G at scale at a single site within the next two years. In addition, 60% of early adopters said 5G helped them increase their operational efficiency while 43% said it made their networks more flexible.

The same survey indicated that many companies (64%) plan to adopt 5G-based edge computing services within the next three years and more than a third of the industrial organizations surveyed said they prefer to deploy 5G private networks, especially those operating in the high-tech, aerospace and defense sectors.

Like commercial networks, 5G private networks consist of three main elements—user equipment, the radio access network (RAN), and the core network—and they typically leverage MEC. No surprises here but industrial devices differ greatly from commercial ones, creating new challenges for design engineers.

Key aspects to consider during design validation

Industrial devices vary widely. They span sensors, adapters, and remote I/Os used in process control and monitoring applications, smart glasses for augmented reality (AR) and asset management use cases, and drones for remote access and maintenance. Then there are device types like network bridges, routers, and network gateways for motion control, mobile robots and other applications.

These devices present unique characteristics. For example, IP67 sensors, which are used in harsh industrial environments for closed-loop process control and monitoring or to issue alerts when certain events occur, have a robust enclosure. Some of these sensors are even filled with epoxy resins or other insulating liquid compounds to protect the electronic components inside them, and they have limited or no external interfaces.

Another use case, 5G smart glasses, support technicians when commissioning smart field devices, performing maintenance tasks, and visualizing process data. They have high uplink (UL)/downlink (DL) traffic and low latency requirements, are battery-operated, and need to be able to operate in special conditions like explosive environments.

Drones also have high availability and low latency requirements, as well as real-time positioning and high uplink bandwidth needs. A 5G-enabled drone can effectively monitor large and distant areas, but that requires high-performance communications. The sensors on the drone send data to the user continuously. Therefore, high availability, security, low latency, real-time positioning information and time-synchronization services are critical aspects for drone control.

Next, 5G wireless routers, which serve automated guided vehicles (AGV) and other mobile machines and local area network (LAN) switches for connected devices, present unique challenges to engineers. A mobile machine featuring a traffic management system and other IT functions will use IP-based traffic as well as industrial Ethernet protocols such as PROFINET.

These devices are also becoming increasingly complex. Future robotic arms, for example, will embed multiple devices to increase position accuracy and versatility. And they will connect directly to a 5G base station or go through a gateway, as shown in Figure 2.

Figure 2 A robotic arm with embedded cellular devices connects to a 5G base station directly or through a gateway. Source: 5G-ACIA

Cellular module integration into an industrial device like a robotic arm is particularly challenging. Validating device connectivity, reliability, and stability as well as the functionality and performance of the device are critical because downtime and errors in manufacturing operations can be costly.

You need to consider all the validation and conformance testing typically needed for a consumer cellular device. In addition, new considerations emerge because almost all industrial devices need to operate in a ruggedized environment.

Once chips and other components are assembled into systems, conformance testing may be required to ensure consistency with the standard and interoperability after the device is deployed. While not prevalent yet, a few private network operators that provide managed services may want to ensure industrial device quality with supplemental tests.

Engineers can perform conformance tests using a network emulator like Keysight’s UXM 5G platform. Performing rigorous testing across the protocol and RF/radio resource management (RRM) domains is essential to achieve certification.

Figure 3 A network emulation platform covers protocol, RF, functional and performance aspects and can scale from benchtop R&D to full-rack acceptance test. Source: Keysight

Data security is also paramount for smart factories, requiring an end-to-end life cycle strategy. This strategy starts and ends with the devices. They are the first and last line of defense against cyberattacks. With new vulnerabilities emerging constantly and new builds introducing new or unknown issues, thoroughly assessing the security of industrial devices is a step that should not be overlooked in the design validation process.

Also, 5G leverages ultra-reliable and low-latency communication (URLLC) capabilities to meet TSN-related quality of service (QoS) requirements. The demand for time synchronization between the devices in a 5G private network, guaranteed latency, and no congestion loss drives the need for testing devices for TSN and URLLC.

The need for channel modeling

In addition to taking into consideration the specificities of industrial devices, understanding the characteristics of the industrial network environment is important to ensure efficient use of a 5G private network. As shown in Figure 4, factories use a lot of metal and high-speed rotating machines; users and physical objects also obstruct the radio frequencies. These aspects degrade the signals by causing reflections, refractions, and fading.

Figure 4 A typical factory environment blocks and reflects RF signals. Source: Keysight

Signal propagation models that represent the factory environment are essential to validate the performance of industrial devices. A typical factory environment includes high vertical spaces and metallic surfaces and is highly dynamic in impacting the radio channel. The model replicating this environment needs to include large elevation spreads, multipaths, reflections, large angular and long delay spreads, and dynamic propagation conditions.

The antennas on an industrial device receive multiple signals from the base station, including direct signals and signals that reflect off of it. Creating a channel model by isolating the device and the base station is the first step in the modeling process. The channel model estimates the delay between multiple paths and the angles at which the base station transmits and receives the signals by taking into consideration propagation factors like path loss.

Once the model is ready, a channel emulator can replay the model to validate the performance of the device in those conditions, as shown in Figure 5. Next, a network emulator can replace the base station to create repeatable signals with a fixed set of parameters. The device needs to receive the signals under excellent conditions though. Isolating the device in a test chamber creates the repeatable environment needed for evaluating the device.

Figure 5 A channel emulator helps test a device to meet the requirements of a challenging factory environment. Source: Keysight

Validation and conformance of industrial devices

The lifecycle of a 5G private network spans multiple stages. The validation and conformance of industrial devices is only the first step, but a critical one. The unique capabilities of industrial devices, their integration requirements, and growing complexity create many challenges that drive the need for testing and channel modeling.

Currently, only a few industrial devices support 5G, partially because cellular technology is new to many industrial device makers. That poses a challenge the industry needs to solve for its digital transformation. Solutions exist to address that challenge. Keysight’s network and channel emulation solutions, for example, support multiple technologies, including LTE and 5G, so engineers can validate the design for 5G industrial device and achieve certification, if desired. A comprehensive test and validation platform also helps save time and cost in the process.

Learn more about network and channel emulation on the web pages 5G network emulation solutions and 5G channel emulation solutions.

Engineers can also learn more about 5G private networks by listening to the latest episode of the All Things 5G podcast.

This article was originally published on EDN.

Jessy Cavazos is part of Keysight’s Industry Solutions Marketing team.

Leave a comment